The major objective of the proposed research is to identify the risk factors that contribute to developing bulimia with the ultimate goal being the formulation of effective prevention strategies. 1116 high school females, in grades 9, 10 and 11 will be surveyed in a cross-sectional, longitudinal design for information on a) bulimic symptomatology, b) weight control practices, c) self concept, d) psychiatric symptomatology (particularly depressive symptoms), e) family environment and f) peer group beliefs and behavior regarding weight control practices. Measurements will take place at 2 points in time separated by 11 mos. The 5 independent variables (b-f) will be combined in a series of multiple regression and discriminant function analyses to predict the amount of bulimic symptomatology (a) at T1 and to predict changes in symptomatology from T1 to T2. Special attention will be paid to predicting which females with a subclinical pattern of symptomatology at T1 become more or less symptomatic at T2. Data collection for T1 has been completed and collection is underway for T2. Funding is being requested to analyze the data from T2, to perform the major longitudinal data analyes using the data sets from T1 and T2, and to design a prevention program/curriculum to be adopted in secondary school curricula. Specific questions to be addressed include: 1) What is the prevalence and incidence of bulimic symptomatology in the high school female population?, 2) What variables best predict increases or decreases in symptomatology from T1 to T2 including weight histories, self concept differences psychiatric symptoms and family variables. Knowledge gained from this research will be used to formulate a prevention curriculum specifying who to target, when to intervene and the content area best suited to each particular target group.